کنفرانس مهندسی برق ایران

صفحه اصلی / سی و سومین کنفرانس بین المللی مهندسی برق

Iranian stock market fluctuations: from social news to forecasting models

نویسندگان :

Maryam Sharifinia¹ Farzaneh Ghayour Baghbani²

1- دانشگاه علم و صنعت ایران 2- دانشگاه علم و صنعت ایران

کلمات کلیدی :

stock price prediction،Sentiment analysis،trading data،stock price change classification،news

چکیده :

This research investigates the impact of news on stock price changes in the Tehran Stock Exchange. Due to the lack of Persian news analysis models for the capital market, news items were translated into English and analyzed using the FinBERT model. Market data was obtained from the Tehran Stock Exchange website and stored after price adjustment. Based on the deposit report, one stock was selected as a representative from each industry, and a separate model was trained for each stock. In the second phase, price changes were categorized into five groups and labeled. To improve accuracy, parameters such as changes in trading volume and the overall index were added to the model to achieve more accurate price change predictions. Finally, the challenges associated with using English models and the need to develop dedicated Persian models for capital market analysis are discussed.